[RTTY] Propagation On 60m in RTTY Contests

[Kok Chen]

On Oct 27, 2009, at 2:17 PM, Thomas F. Giella NZ4O wrote:

> The 160 meter RTTY distortion that you describe is very minor and also
> occurs on CW and SSB.

Of course it occurs for every other mode.  I have not said otherwise.

I just wanted to point out that RTTY, with a 22ms symbol rate, is much  
more susceptible (i.e., even loud signals print with lots of errors)  
to multipath delays of the order of 10ms than a digital modulation  
mode (such as DominoEX or Throb) that is specifically designed to take  
long multipath into consideration.

As Murray ZL1BPU has observed on an NVIS channel, multipath effects  
are often bad at certain times of the day.  When multipath is bad, the  
symptom is that the error rate no longer decrease as you increase SNR  
(e.g., transmit power).  Most of us have seen cases where we can't I  
print a strong signal.

A common problem with digital modulation modes is the ISI (intersymbol  
interference) in the presence of multipath.  RTTY is not really good  
for the case when the multipath allows RTTY symbols to overlap   
significantly.

I am all in favor of more 160m RTTY activity since we can more easily  
study ways to improve demodulators, and perhaps even validate or  
invalidate the ITU models for the lower bands -- the original IEEE  
paper by Watterson that the ITU specs is based on had looked at  
signals at 9.26 MHz and 5.86 MHz.  The 9 MHz signal showed multipath  
delays of around 40 us (E layer) and weaker paths in the 300 us and  
1ms (one and two F layer rays).

The 5.8 MHz signal had a significantly longer multipath delay of about  
450 us, with weaker components at 750 us and 1ms, all of which are  
listed as "unknown" propagation modes.

The fact that RTTY is hard to decode on 160m will give an edge to  
those who have better decoders -- that alone could spur the  
development of better demodulators.

73
Chen, W7AY


By the way, speaking of multipath, I was digging through my email  
exchanges with Alex VE3NEA back in early 2005.  We were trading sound  
files that we had recorded of hard-to-print signals to challenge each  
other's demodulator (in case you are not aware, Alex had an RTTY  
demodulator which he never bothered to publish -- he is always  
listening in RTTY contests, but I never heard him transmit during any  
of them).

I had sent him a recording of W4JO in a QSO with VK2IMM.  The W4 was  
very loud, but still dropping print now and then.  I had sent a time  
domain plot to Alex and in his reply, Alex had pointed to a sudden  
rise in echo that lasted for a little less than 1 second.  I had not  
noticed the echos before that, since it was *more* than 22ms (the  
largest delay was 40ms !).  At the peak, the echo from the start bit  
of a diddle (LTRS) character was easily visible at about -15 dB down  
from the actual start bit.  This appeared on the onset of the signal  
selectively fading, but that may be entirely uncorrelated.

In a followup e-mail, Alex had told me that he was getting better copy  
of that echo signal by turning off the ATC (in spite of the fact that  
there was selective fading).

Alex went on to have fun with the SoftRock and wideband CW decoders  
(culminating in the CW Skimmer), but we might be able to get him  
interested again in studying RTTY if we present him with the 160m  
challenge :-).